Papers Containing Tag(s): 'Maximum Likelihood Estimation'
The following papers contain search terms that you selected. From the papers listed below, you can navigate to the PDF, the profile page for that working paper, or see all the working papers written by an author. You can also explore tags, keywords, and authors that occur frequently within these papers.
See Working Papers by Tag(s), Keywords(s), Author(s), or Search Text
Click here to search again
Frequently Occurring Concepts within this Search
Authors
No authors occur more than twice in this search.
Viewing papers 1 through 10 of 11
-
Working PaperTechnology-Driven Market Concentration through Idea Allocation
December 2025
Working Paper Number:
CES-25-78
Using a newly-created measure of technology novelty, this paper identifies periods with and without technology breakthroughs from the 1980s to the 2020s in the US. It is found that market concentration decreases at the advent of revolutionary technologies. We establish a theory addressing inventors' decisions to establish new firms or join incumbents of selected sizes, yielding two key predictions: (1) A higher share of inventors opt for new firms during periods of heightened technology novelty. (2). There is positive assortative matching between idea quality and firm size if inventors join incumbents. Both predictions align with empirical findings and collectively contribute to a reduction in market concentration when groundbreaking technologies occur. Quantitative analysis shows the overall slowdown in technological breakthroughs can capture 95.9% of the rising trend in market concentration and the correlation between the model-generated and the actual detrended market concentration is 0.910.View Full Paper PDF
-
Working PaperInvestigating the Effect of Innovation Activities of Firms on Innovation Performance: Does Firm Size Matter?
January 2025
Working Paper Number:
CES-25-04
Understanding the relationship between a firm's innovation activities and its performance has been of great interest to management scholars. While the literature on innovation activities is vast, there is a dearth of studies investigating the effect of key innovation activities of the firm on innovation outcomes in a single study, and whether their effects are dependent on the nature of firms, specifically firm size. Drawing from a longitudinal dataset from the Business Research & Development and Innovation Survey (BRDIS), and informed by contingency theory and resource orchestration theory, we examine the relationship between a firm's innovation activities - including its Research & Development (R&D) investment, securing patents, collaborative R&D, R&D toward new business areas, and grants for R&D - and its product innovation and process innovation. We also investigate whether these relationships are contingent on firm size. Consistent with contingency theory, we find a significant difference between large firms and small firms regarding how they enhance product innovation and process innovation. Large firms can improve product innovation by securing patents through applications and issuances, coupled with active participation in collaborative R&D efforts. Conversely, smaller firms concentrate their efforts on the number of patents applied for, directing R&D efforts toward new business areas, and often leveraging grants for R&D efforts. To achieve process innovation, a similar dichotomy emerges. Larger firms demonstrate a commitment to securing patents, engage in R&D efforts tailored to new business areas, and actively collaborate with external entities on R&D efforts. In contrast, smaller firms primarily focus on securing patents and channel their R&D efforts toward new business pursuits. This nuanced exploration highlights the varied strategies employed by large and small firms in navigating the intricate landscape of both product and process innovation. The results shed light on specific innovation activities as antecedents of innovation outcomes and demonstrate how the effectiveness of such assets is contingent upon firm size.View Full Paper PDF
-
Working PaperWhere to Build Affordable Housing? Evaluating the Tradeoffs of Location
December 2023
Working Paper Number:
CES-23-62R
How does the location of affordable housing affect tenant welfare, the distribution of assistance, and broader societal objectives such as racial integration? Using administrative data on tenants of units funded by the Low-Income Housing Tax Credit (LIHTC), we first show that characteristics such as race and proxies for need vary widely across neighborhoods. Despite fixed eligibility requirements, LIHTC developments in more opportunity-rich neighborhoods house tenants who are higher income, more educated, and far less likely to be Black. To quantify the welfare implications, we build a residential choice model in which households choose from both market-rate and affordable housing options, where the latter must be rationed. While building affordable housing in higher-opportunity neighborhoods costs more, it also increases household welfare and reduces city-wide segregation. The gains in household welfare, however, accrue to more moderate-need, non-Black/Hispanic households at the expense of other households. This change in the distribution of assistance is primarily due to a 'crowding out' effect: households that only apply for assistance in higher-opportunity neighborhoods crowd out those willing to apply regardless of location. Finally, other policy levers'such as lowering the income limits used for means-testing'have only limited effects relative to the choice of location.View Full Paper PDF
-
Working PaperLabor Market Segmentation and the Distribution of Income: New Evidence from Internal Census Bureau Data
August 2023
Working Paper Number:
CES-23-41
In this paper, we present new findings that validate earlier literature on the apparent segmentation of the US earnings distribution. Previous contributions posited that the observed distribution of earnings combined two or three distinct signals and was thus appropriately modeled as a finite mixture of distributions. Furthermore, each component in the mixture appeared to have distinct distributional features hinting at qualitatively distinct generating mechanisms behind each component, providing strong evidence for some form of labor market segmentation. This paper presents new findings that support these earlier conclusions using internal CPS ASEC data spanning a much longer study period from 1974 to 2016. The restricted-access internal data is not subject to the same level of top-coding as the public-use data that earlier contributions to the literature were based on. The evolution of the mixture components provides new insights about changes in the earnings distribution including earnings inequality. In addition, we correlate component membership with worker type to provide a tacit link to various theoretical explanations for labor market segmentation, while solving the problem of assigning observations to labor market segments a priori.View Full Paper PDF
-
Working PaperHeavy Tailed, but not Zipf: Firm and Establishment Size in the U.S.
July 2021
Working Paper Number:
CES-21-15
Heavy tails play an important role in modern macroeconomics and international economics. Previous work often assumes a Pareto distribution for firm size, typically with a shape parameter approaching Zipf's law. This convenient approximation has dramatic consequences for the importance of large firms in the economy. But we show that a lognormal distribution, or better yet, a convolution of a lognormal and a non-Zipf Pareto distribution, provides a better description of the U.S. economy, using confidential Census Bureau data. These findings hold even far in the upper tail and suggest heterogeneous firm models should more systematically explore deviations from Zipf's law.View Full Paper PDF
-
Working PaperThe Energy Efficiency Gap and Energy Price Responsiveness in Food Processing
June 2020
Working Paper Number:
CES-20-18
This paper estimates stochastic frontier energy demand functions with non-public, plant-level data from the U.S. Census Bureau to measure the energy efficiency gap and energy price elasticities in the food processing industry. The estimates are for electricity and fuel use in 4 food processing sectors, based on the disaggregation of this industry used by the National Energy Modeling System Industrial Demand Module. The estimated demand functions control for plant inputs and output, energy prices, and other observables including 6-digit NAICS industry designations. Own price elasticities range from 0.6 to -0.9 with little evidence of fuel/electricity substitution. The magnitude of the efficiency estimates is sensitive to the assumptions but consistently reveal that few plants achieve 100% efficiency. Defining a 'practical level of energy efficiency' as the 95th percentile of the efficiency distributions and averaging across all the models result in a ~20% efficiency gap. However, most of the potential reductions in energy use from closing this efficiency gap are from plants that are 'low hanging fruit'; 13% of the 20% potential reduction in the efficiency gap can be obtained by bringing the lower half of the efficiency distribution up to just the median level of observed performance. New plants do exhibit higher energy efficiency than existing plants which is statistically significant, but the difference is small for most of the industry; ranging from a low of 0.4% to a high of 5.7%.View Full Paper PDF
-
Working PaperRelative Effectiveness of Energy Efficiency Programs versus Market Based Climate Policies in the Chemical Industry
April 2018
Working Paper Number:
CES-18-16
This paper addresses the relative effectiveness of market vs program based climate policies. We compute the carbon price resulting in an equivalent reduction in energy from programs that eliminate the efficiency gap. A reduced-form stochastic frontier energy demand analysis of plant level electricity and fuel data, from energy-intensive chemical sectors, jointly estimates the distribution of energy efficiency and underlying price elasticities. The analysis controls for plant level price endogeneity and heterogeneity to obtain a decomposition of efficiency into persistent (PE) and time-varying (TVE) components. Total inefficiency is relatively small and price elasticities are relatively high. If all plants performed at the 90th percentile of their efficiency distribution, the reduction in energy is between 4% and 13%. A modest carbon price of between $9.48/ton and $14.01/ton CO2 would achieve reductions in energy use equivalent to all manufacturing plants making improvements to close the efficiency gap.View Full Paper PDF
-
Working PaperRanking Firms Using Revealed Preference
January 2017
Working Paper Number:
CES-17-61
This paper estimates workers' preferences for firms by studying the structure of employer-toemployer transitions in U.S. administrative data. The paper uses a tool from numerical linear algebra to measure the central tendency of worker flows, which is closely related to the ranking of firms revealed by workers' choices. There is evidence for compensating differential when workers systematically move to lower-paying firms in a way that cannot be accounted for by layoffs or differences in recruiting intensity. The estimates suggest that compensating differentials account for over half of the firm component of the variance of earnings.View Full Paper PDF
-
Working PaperHOW IMPORTANT ARE SECTORAL SHOCKS
September 2014
Working Paper Number:
CES-14-31
I quantify the contribution of sectoral shocks to business cycle fluctuations in aggregate output. I develop a multi-industry general equilibrium model in which each industry employs the material and capital goods produced by other sectors, and then estimate this model using data on U.S. industries sales, output prices, and input choices. Maximum likelihood estimates indicate that industry-specific shocks account for nearly two-thirds of the volatility of aggregate output, substantially larger than previously assessed. Identification of the relative importance of industry-specific shocks comes primarily from data on industries intermediate input purchases, data that earlier estimations of multi-industry models have ignored.View Full Paper PDF
-
Working PaperEstimating Record Linkage False Match Rate for the Person Identification Validation System
July 2014
Working Paper Number:
carra-2014-02
The Census Bureau Person Identification Validation System (PVS) assigns unique person identifiers to federal, commercial, census, and survey data to facilitate linkages across files. PVS uses probabilistic matching to assign a unique Census Bureau identifier for each person. This paper presents a method to measure the false match rate in PVS following the approach of Belin and Rubin (1995). The Belin and Rubin methodology requires truth data to estimate a mixture model. The parameters from the mixture model are used to obtain point estimates of the false match rate for each of the PVS search modules. The truth data requirement is satisfied by the unique access the Census Bureau has to high quality name, date of birth, address and Social Security (SSN) data. Truth data are quickly created for the Belin and Rubin model and do not involve a clerical review process. These truth data are used to create estimates for the Belin and Rubin parameters, making the approach more feasible. Both observed and modeled false match rates are computed for all search modules in federal administrative records data and commercial data.View Full Paper PDF
